Regional Mapping and Reservoir Analysis of the Middle Devonian Marcellus Shale in the Appalachian Basin

Home , Huntersville Chert, Interbedding, Isopach map, Needmore Shale

Graduate Theses, Dissertations, and Problem Reports

2014

Jared VanMeter West Virginia University

Follow this and additional works at: https://researchrepository.wvu.edu/etd

Recommended Citation VanMeter, Jared, "Regional Mapping and Reservoir Analysis of the Middle Devonian Marcellus Shale in the Appalachian Basin" (2014). Graduate Theses, Dissertations, and Problem Reports. 634. https://researchrepository.wvu.edu/etd/634

This Thesis is protected by copyright and/or related rights. It has been brought to you by the The Research Repository @ WVU with permission from the rights-holder(s). You are free to use this Thesis in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you must obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/ or on the work itself. This Thesis has been accepted for inclusion in WVU Graduate Theses, Dissertations, and Problem Reports collection by an authorized administrator of The Research Repository @ WVU. For more information, please contact [email protected].

Regional Mapping and Reservoir Analysis of the Middle Devonian Marcellus Shale in the Appalachian Basin

Jared VanMeter

Thesis submitted to the Eberly College of Arts and Sciences at West Virginia University

in partial fulfillment of the requirements for the degree of

Master of Science in Geology

Richard Smosna, Ph.D., Chair Kathy Bruner, Ph.D. Tim Carr, Ph.D.

Department of Geology and Geography

Morgantown, West Virginia 2013

ABSTRACT

Regional Mapping and Reservoir Analysis of the Middle Devonian Marcellus Shale in the Appalachian Basin

Jared VanMeter

The main purpose of this investigation is to define the distribution of organic-rich facies of the Middle Devonian Marcellus Shale in New York, Ohio, Pennsylvania, and West Virginia. The analysis is based on well-log data, primarily gamma-ray (the most common and best calibrated) and bulk-density logs (where available). Detailed log analysis has been performed to normalize the logs and define key indicators of reservoir quality. Stratigraphic correlations have been conducted to trace key formation across the study area. The following maps were generated over the study area.  Isopach maps of the Mahantango Shale, Marcellus Shale, Oatka Creek Member, Cherry Valley and Union Springs Member. These maps show the stratigraphic thicknesses of the various formations and members.  Net thickness maps of the Marcellus Shale where the gamma-ray < 100 API and gamma-ray is between 100-180 API. The gamma-ray < 100 API map shows the thickness of the various limestone intervals within the Marcellus. The paleography of these intervals represents carbonate shoals along the basin margin (north and west) and over the peripheral bulge. The gamma-ray map between 100-180 API shows the net thickness of calcareous shale and gray shale. These intervals represent the shallow muddy sea above the thermocline.  Net thickness maps of the Marcellus Shale where the gamma-ray >180 API, > 200 API, > 250 API, and > 300 API. These maps show the thickness and distribution of shale with different organic-richness within the Marcellus. The paleography of these shale intervals was the deep basin below the thermocline. These maps show the location of the better reservoir.  Average gamma-ray over the Marcellus interval. This map shows the average gamma-ray value for the Marcellus Shale across the basin. The map can be used as an indication of the highest average organic-richness of the Marcellus Shale, and is best used in conjunction with the net thickness maps by comparing the thickest portion of the various maps and where they may overlap regionally.  Net thickness maps of the Marcellus Shale where the bulk-density < 2.55 g/cc, < 2.4 g/cc, and < 2.35 g/cc. These maps show various reservoir quality grades of organic-richness (where lower density equals higher quality).  Isopach Maps of stratigraphic sequences and their systems tracts in the Marcellus Shale. Thin Transgressive Systems Tracts equal a condensed section, whereas thick Regressive Systems Tracts equal a major clastic influx. These observations and others portrayed on this new series of maps provide a better understanding of the exploratory development of the Marcellus Shale in the Appalachian Basin. By using all of these maps in conjunction, the best target areas for oil and gas exploration can be identified and exploited. The sequence stratigraphic maps can be used for regional correlations and to develop target zones within the Marcellus Shale.

ACKNOWLEDGMENTS

I would like to express appreciation and thanks to my advisor Dr. Richard Smosna, you have been a tremendous mentor for me and an inspiration in my work. I would like to thank you for encouraging my research and for allowing me to grow as a Geologist. Your advice on my research as well as on my career has been priceless. I would also like to thank my committee members, Dr. Kathy Bruner, and Dr. Tim Carr. Thank you both for listening to my defense, and for your brilliant comments and suggestions on my thesis. I would especially like to thank all the faculty and staff of the Geology Department at WVU. You all have helped mold me into the professional that I am today.

I would like to give a special thank you to the National Energy Technology Laboratories in Morgantown WV for funding my research and providing the data therein. I am very thankful that they allowed me to put this research into thesis format.

I would also like to thank all of my friends who supported me in writing, and encouraged me to strive towards finishing.

A special thanks to my family. Words cannot express how grateful I am to my mother-in law, Johnnie and my late father-in-law Fred. Fred, you may not be able to see me graduate like you wanted, but you will always be there in spirit. I would like to thank my mother and father,

Kristi and Guy for all of the sacrifices that you’ve made on my behalf. You both were my guiding light and I couldn’t have done it without you. At the end I would like express appreciation to my beloved wife Katie VanMeter who was there from beginning to end through the thick and thin. Who spent her evenings helping me search for data and digitizing logs on my behalf. There is no level of thanks that could express all of my gratitude for this!

i TABLE OF CONTENTS

ACKNOWLEDGMENTS……………………………………………………………… I

LIST OF FIGURES……………………………………………………………………… IV

INTRODUCTION…..…………………………………………………………………… 1

MIDDLE DEVONIAN STRATIGRAPHY OF THE APPALACHIAN BASIN……… 3

DEPOSITIONAL SETTING OF THE MARCELLUS AND MAHANTANGO

FROMATIONS………………………………………………………………………… 15

AREA OF INVESTIGATION…………………………………………………………… 21

METHODS OF INVESTIGATION…….……………………………………………… 23

ISOPACH MAPS………………………………………………………………………… 28

MARCELLUS SHALE ISOPACH.…………………………………………………… 28

UNION SPRINGS MEMBER ISOPACH……………………………………… 33

CHERRY VALLEY MEMBER ISOPACH…………………………………… 35

OATKA CREEK MEMBER ISOPACH………………………………………… 37

MAHANTANGO FORMATION ISOPACH…………………………………… 39

STRATIGRAPHIC CROSS SECTIONS...……………………………………………… 41

A-A’ CROSS SECTION………………………………………………………… 43

B-B’ CROSS SECTION………….……………………………………………… 45

C-C’ CROSS SECTION………………………………………………………… 47

D-D’ CROSS SECTION………………………………………………………… 49

E-E’ CROSS SECTION…..……………………………………………………… 51

F-F’ CROSS SECTION………………………………………………………… 53

G-G’ CROSS SECTION..……………………………………………………… 55

ii H-H’ CROSS SECTION..……………………………………………………… 57

I-I’ CROSS SECTION.………………………………………………………… 59

J-J’ CROSS SECTION…………………………………………………………… 62

PETROLEUM RESERVOIR MAPS …………………………………………………… 65

GAMMA RAY ISOPACH MAPS……………………………………………… 65

BULK-DENSITY ISOPACH MAPS…..………………………………………… 79

SEQUENCE STRATIGRAPHY………………………………………………………… 85

TRANSGRESSIVE REGRESSIVE SEQUENCE 1…………………………… 89

TRANSGRESSIVE REGRESSIVE SEQUENCE 2 ...………………………… 95

TRANSGRESSIVE REGRESSIVE SEQUENCE 3..…………………………… 101

DISCUSSION…………………………………………………………………………… 107

SEDIMENTOLOGY, STRATIGRAPHY, AND DEPOSITIONAL SETTING.. 107

PETROLEUM ISOPACH MAPS……………………………………………… 112

COMPARISON OF GAMMA-RAY NAD BULK-DENSITY ISOPACH MAPS 118

SEQUENCE STRATIGRAPHY………………………………………………… 121

CONCLUSIONS...……………………………………………………………………… 125

REFERENCES.………………………………………………………………………… 128

iii LIST OF FIGURES

Figure 1. Isopach map of the Hamilton Group……………………………………………. 6

Figure 2A. Hamilton Group Stratigraphy for New York and northern Pennsylvania……. 9

Figure 2B. Hamilton Group Stratigraphy for southern Pennsylvania and West Virginia… 9

Figure 3. Isopach map of Mahantango Shale……………………………………………. 11

Figure 4. Isopach map of Tully Limestone………………………………………………. 14

Figure 5. Study Area……………………………………………………………………. 22

Figure 6. Location of data control points……………………………………………….. 24

Figure 7. Index line map for cross sections with total well spots……………………….. 26

Figure 8. Isopach map of Marcellus Shale………………………………………………. 29

Figure 9. Growth fault Rom Trough map………………………………………………… 30

Figure 10. Map of major basement faults and lineaments………………………………… 32

Figure 11. Isopach map of the Union Springs Member of the Marcellus Shale………….. 34

Figure 12. Isopach map of the Cherry Valley Member of the Marcellus Shale………… 36

Figure 13. Isopach map of the Oatka Creek Member of the Marcellus Shale…………….. 38

Figure 14. Isopach map of the Mahantango Shale………………………………………... 40

Figure 15. Regional map showing the locations of all cross sections……………………. 42

Figure 16. A-A’ Cross section…………………………………………………………….. 44

Figure 17. B-B’ Cross section…………………………………………………………….. 46

Figure 18. C-C’ Cross section…………………………………………………………….. 48

Figure 19. D-D’ Cross section…………………………………………………………….. 50

Figure 20. E-E’ Cross section…………………………………………….……………….. 52

Figure 21. F-F’ Cross section…………………………………………………………….. 54

Figure 22. G-G’ Cross section…………………………………………………………….. 56

Figure 23. H-H’ Cross section…………………………………………………………….. 58

iv Figure 24. I-I’ Cross section………………………………………………………………. 60

Figure 25. J-J’ Cross section………………………………………………………………. 63

Figure 26. Marcellus net thickness less than 100 API…………………………………… 66

Figure 27. Marcellus net thickness between 100 and 180 API…………………………… 68

Figure 28. Marcellus net thickness greater than 180 API………………………………… 70

Figure 29. Marcellus net thickness greater than 200 API………………………………… 72

Figure 30. Marcellus net thickness greater than 250 API………………………………… 74

Figure 31. Marcellus net thickness greater than 300 API………………………………… 76

Figure 32. Marcellus average gamma-ray API Map……………………………………… 78

Figure 33. Marcellus bulk-density net thickness map of feet less than 2.55 g/cc………….80

Figure 34. Marcellus bulk-density net thickness map of feet less than 2.4 g/cc………… 82

Figure 35. Marcellus bulk-density net thickness map of feet less than 2.35 g/cc………… 84

Figure 36. Sequence Stratigraphic type log……………………………………………… 86

Figure 37. Location of sequence stratigraphic cross section……………………………… 87

Figure 38. B-B’ sequence stratigraphic cross section …………………………………… 88

Figure 39. Isopach map of the first Transgressive Regressive Sequence…………………. 90

Figure 40. Isopach map of the first Transgressive Systems Tract………………………… 92

Figure 41. Isopach map of the first Regressive Systems Tract…………………………… 94

Figure 42. Isopach map of the second Transgressive Regressive Sequence……………. 96

Figure 43. Isopach map of the second Transgressive Systems Tract…………………… 98

Figure 44. Isopach map of the second Regressive Systems Tract………………………… 100

Figure 45. Isopach map of the third Transgressive Regressive Sequence………………… 102

Figure 46. Isopach of the third Transgressive Systems Tract…………………………… 104

Figure 47. Isopach map of the third Regressive Systems Tract………………………… 106

Figure 48. Paleogeographic Map………………………………………………………….. 109

v Figure 49. Cross plot of RhoB vs TOC (From Laughrey 2009)………………………… 116

Figure 50. Cross Plot of RhoB vs TOC% (From Cluff 2009)…………………………… 117

Figure 51. Cross Plot of 2.5 g/cc RhoB vs gamma-ray greater than 180 API………….. 120

INTRODUCTION

The Devonian Marcellus Shale of the Appalachian Basin has been producing gas since the late 1800s, but only recently has it become a giant play. In 2005, Range Resources, using new drilling and fracturing techniques, began producing gas from a Marcellus Shale well in

Washington County, Pennsylvania (Perry and Wickstrom 2010). Now the Marcellus Shale is the largest producer of unconventional shale gas in the United States. During the Eastern Gas Shale

Project in the 1970s, it was estimated that the Marcellus contains 295 trillion cubic feet of gas in place throughout the basin (Harper 2008). Today it is estimated the Marcellus contains 500 trillion cubic feet of gas in place, with 50 trillion cubic feet of recoverable gas (Harper 2008).

Geologists in both academia and industry have mapped the Marcellus Shale extensively.

However, there are many different methods and criteria for mapping, so that each study shows something different. Most authors identify the Marcellus Shale in the subsurface based upon a minimum gamma-ray value of 200 API units, which can typically be correlated to a TOC content of 5% or more (Myers 2008, Lash and Engelder 2009, Boyce 2010, Zagorski and Wrightstone

2012). Other authors, however, identify the shale by a gamma-ray value as low as 175-180 API units (Piotroski and Harper 1979, Hill et al. no date). Still other authors identify the shale by a minimum gamma-ray value of 230 API units or higher (Cliff Minerals Inc. 1979, Hill et al. 2002,

Zaengle 2008, Avery and Lewis no date). The purpose of this thesis research is to better understand the distribution of the organic-rich facies in the Marcellus Shale plus the overlying

Mahantango Formation by using common well logging techniques and to clearly define the criteria for mapping these rocks.

1 A second aspect of this present study is to correlate typical bulk-density values to the organic-richness of shale. Cliff Minerals Inc. (1979) showed that black shale can have bulk- density values as low as 2.20-2.30 gm/cc. This low density correlates to a TOC of 10% - 12%

(Gottschling 2007, Engelder 2008) and up to 20% (Smith and Leone 2009). Low bulk-density reading is thought to represent the best, most organic-rich portion of the shale. Boyce (2010) used a much higher density cut-off value of 2.55gm/cc to represent organic shale which correlates to ~6-7% TOC.

Another aspect of this thesis is to identify, correlate, and map stratigraphic sequences across the Appalachian Basin. Previous sequence stratigraphic studies of the Marcellus Shale have identified small-scale stratigraphic changes that may help in predicting reservoir plays in areas with a low data spread. Mapping such sequence stratigraphic features may also help in selecting local target zones for directional drilling within the formation. The prominent gamma- ray signatures of these features can provide predictability in steering the drill bit in a lateral wellbore. Lash (2010) used the transgressive-regressive (T-R) sequence model in order to divide the Marcellus Shale into easily definable sequences defined by the rise and subsequent fall of relative sea level (Catuneanu 2006). These systems tracts of transgression and regression can be easily mapped from outcrop studies and gamma-ray logs. The maximum flooding surface is represented by the highest peak of the gamma-ray log in formation. The maximum regression is marked by the lowest trough of the gamma-ray log in the formation (Embry 2002, Catuneanu

2006, Lash 2010, Lash and Blood 2011). Mapping these key sequence stratigraphic surfaces should bring about a general understanding of the sequence of sedimentary events in the formation of the Marcellus Shale.

2 MIDDLE DEVONIAN STRATIGRAPHY OF THE APPALACHIAN BASIN

The Middle Devonian Marcellus Shale is located in the northern part of the Appalachian

Basin, primarily within the Allegheny Plateau province. The formation lies within the following states: southern New York, much of Pennsylvania, eastern Ohio, western Maryland, most of

West Virginia, northern and western Virginia, and eastern Kentucky. The formation also extends across Lake Erie into southern Ontario. This area measures 50,000 square miles (Figure

3).

The Marcellus is the lower of two formations in the Hamilton Group. The Tully

Limestone and above and the Onondaga Limestone, Huntersville Chert, or Needmore Shale below bound the group stratigraphically. Wherever these adjacent formations are present, the

Hamilton Group can be delineated clearly in the subsurface (Reeves and Davies 1937, Dennison and Hasson 1976, Milici and others 2006).

Rocks underlying the Hamilton Group belong to the Onesquethaw Group (Schwietering

1981, Wrightstone 2008). The Onondaga Limestone ranges in thickness from 200 feet in western New York to less than 50 feet in central Pennsylvania (Figure 9.) (Reeves and Davies

1937, Van Tyne 1996). The formation consists primarily of calcarenitic to cherty to argillaceous limestones and minor shale (Brett and Ver Straeten 1994). The depositional environment was a shallow carbonate shelf in the western and northern portions of the basin.

East of the Onondaga Limestone in West Virginia and Pennsylvania lies the age-equivalent

Huntersville Chert. The chert trends through the center of the Appalachian Basin and ranges in thickness from 10-200 feet. The Huntersville Chert grades from impure chert in West Virginia

3 into the cherty Onondaga Limestone in Pennsylvania and New York (Dennison 1961). There are seven main lithotypes within the Huntersville Chert: (1) clean chert, (2) chert with organic material, (3) spicular chert, (4) dolomitic chert, (5) glauconitic chert, (6) silty argillaceous chert, and (7) dolomitic silty argillaceous chert. However all of the chert contains some dolomite and quartz silt (Sherrard and Heald 1984). According to Cecil (2004), the source of the silica in the

Huntersville Chert is due to biotic extraction of silica from seawater as well as aeolian sand and dust. East of the chert in West Virginia lies the Needmore Shale. The Needmore Shale is black or brown shale and interbedded limestone that was deposited in the deepest portion of the basin, the source of terrestrial sediments originating from the newly raised Acadian highlands to the east. The formation ranges in thickness from 100 feet in the far east to 10 feet near the basin center (Reeves and Davies 1937, Hasson and Dennison 1976, Van Tyne. 1996).

The zone of Tioga K-bentonite occurs near the base of the Marcellus, within the equivalent Millboro Shale in southern West Virginia, or in the underlying Needmore Shale or

Huntersville Chert. The Tioga K-bentonite zone is composed of several individual ash deposits characterized by brown tuffaceous shale beds with biotite flakes and feldspar. The middle coarse zone consists of three thin tuffaceous beds that comprise a stratigraphic interval of less than 10 feet. The Tioga middle coarse zone marks the top of the Devonian Ornesquethaw stage. The origin of the ash bed was a volcano thought to be located in central Virginia around the area of

Charlottesville. Due to the instantaneous nature of deposition, the Tioga is a stratigraphic time line dated at 390 ± 0.5 million years old (Dennison and Hasson 1961, Dennison and Textoris

1970, Hasson and Dennison 1987, Ver Straeten 2004).

4 The Marcellus Shale and the Mahantango Formation combine to form the Hamilton

Group (Figure 1). In New York, Pennsylvania, and West Virginia the boundary between the formations can be picked from well logs by a gamma-ray signature over 200 API. The Hamilton

Group ranges from 250ft thick in western New York and Pennsylvania to 2,450 feet thick towards central New York and Pennsylvania (Reeves and Davies 1937).

Figure 1. Isopach map of the Hamilton Group. Contour interval is 100 feet.

The Marcellus Shale consists of radioactive, organic-rich black shale and gray shale with interbeds of calcareous shale and thin limestone (Schwietering 1981, Walker-Milani 2011). The shale contains finely macerated organic detritus which is generally responsible for the dark color of the rock (Roen 1984). The Marcellus is extremely fissile and exhibits “poker chip” fissility in outcrop (Avary 2009). The thinly laminated shale ranges in color from medium to light grey and black. The black shale occurs in two main variations: brownish black and olive black. The color is representative of high levels of iron, organic matter, and the calcite content within the shale (Dennison and Hasson 1976, Roen 1984).

In northern Pennsylvania and southwestern New York, the Marcellus Shale is divided into three members. The lowest is the organic-rich Union Springs Shale Member (Lash 2008,

Wrightstone 2008). The Union Springs ranges in thickness from 200 feet in eastern

Pennsylvania and New York to less than 10 feet in central Pennsylvania and western West

Virginia (Figure 10). Erosion removed the Union Springs in a NE-SW trend in northwestern

Pennsylvania and southwestern New York (Lash 2008). According to Lash (2008), the unconformity developed over a topographic high on the sea floor immediately after deposition of the Union Springs. This topographic high may have been a foreland bulge caused by the crust’s isostatic response to orogenic loading in the eastern highlands. The middle member of the

Marcellus is a calcareous shale and limestone member named the Cherry Valley in New York

(Lash 2008, Wrightstone 2008). The Cherry Valley Member has an average thickness of 10 feet but can be as thick as 100 feet in southern Pennsylvania and eastern West Virginia. The Cherry

7 Valley is generally thought to be stratigraphically equivalent to the Purcell Member in West

Virginia (Dennison and Hasson 1976, Neal 1979). Finally, the Oatka Creek Shale Member of the Marcellus Shale is organic-rich at the base and grades upward into undifferentiated gray shale. The Oatka Creek ranges in thickness from 10 feet in western Pennsylvania and southern

West Virginia to over 100 feet thick in eastern Pennsylvania and New York (Figure 2A-B)

(Dennison and Hasson 1976, Schwietering 1981, Milici et al. 2006, Lash 2007).

Figure 2A. Hamilton Group stratigraphy for New York and northern Pennsylvania.

Figure 2B. Hamilton Group stratigraphy for southern Pennsylvania and West Virginia.

9 To the west, in Ohio the Marcellus is equivalent to the Delaware Limestone. In northern

Virginia and southern West Virginia, the Marcellus Formation becomes equivalent to the lower

Millboro Shale (Milici et al. 2006). In this region, the Mahantango Formation and Tully

Limestone pinch out so that the Upper Devonian Harrell Shale rest directly upon the Marcellus.

This combined interval --Marcellus Shale through Harrell Shale-- is termed the Millboro Shale

(Dennison and Hasson 1976).

The thickness of the Mahantango formation ranges from 1,600 feet in eastern

Pennsylvania, western Maryland, and southeastern New York, and pinches out in western

Pennsylvania and southern West Virginia (Figure 3) (Dennison and Hasson 1976, Hasson and

Dennison 1987, Milici et al. 2006, Harper 2008, Boyce 2010). In western New York and

Pennsylvania, the Mahantango is divided into several formations such as the Skaneateles,

Ludlowville, and Moscow Shale. Directly above the Marcellus in northwestern Pennsylvania and southwestern New York lies the Stafford Limestone Formation (Fig 2A).

Figure 3. Isopach map of the Mahantango Shale. The red solid line represents the main body of the clastic wedge. Contour interval is 100 feet.

11 In West Virginia, the Mahantango Formation is divided into three major members that are predominantly siltstone with some limestone and sandstone interbeds. These change facies into undifferentiated grey shale in north-central West Virginia. The members from bottom to top are the Gander Run Shale, the Chaneysville Siltstone, the Frame Shale, and the Clearville

Siltstone (Hasson and Dennison 1987). (Fig 2B)

The Mahantango is dominated by thickly laminated, very silty shale and grey shale interbedded with siltstone, sandstone, and limestone beds. Lash (2007) recognized many small packages within the Mahantango equivalent of New York showing a coarsening upward trend of organic-rich shale at the base, passing upward into undifferentiated gray shale and finally into siltstone with a cap of limestone or sandstone. There are three major depositional systems within these Mahantango rocks: storm dominated offshore marine mudstone and fine sandstone, storm dominated nearshore marine shelf and shoreface sandstone, and tide-dominated shoreline deposits consisting of variably bedded sandstone, pebbly sandstone and mudstone (Prave et al.

1996). The cyclicity of the Mahantango formation depicts an episodic advance and retreat of a tide-dominated shoreline onto a storm-dominated shelf.

The formation overlying the Hamilton Group is the Tully Limestone. The Tully

Limestone is absent in western Pennsylvania and New York and southern parts of West Virginia, which makes the proper identification of the upper Hamilton Group boundary extremely difficult. In southern West Virginia, the Tully becomes a part of the Millboro Shale or disappears altogether. The Tully Limestone is composed of dark-gray argillaceous limestone.

As it passes eastward, it becomes a calcareous shale in south-central Pennsylvania, western

Maryland, and the eastern panhandle of West Virginia. The limestone is thickest along the axis

12 of anticlines such as the Chestnut Ridge anticline in the center of the basin at 90 feet thick, and decreases in thickness to the west, east, and south from the center portion of the basin. The increased thickness on anticlines could be the cumulative effect of many antithetic faults. The limestone was deposited in a shallow marine setting along the central portion of the

Epicontinental Sea (Figure 4.) (Roen 1984, Schwietering 1981, Hasson and Dennison 1987).

Figure 4. Isopach map of the Tully Limestone. Contour interval is 20 feet.

14 DEPOSITIONAL SETTING OF THE MARCELLUS AND MAHANTANGO

The Devonian was a time of eustatic sea-level rise (Ettensohn 2008). During the Acadian orogeny, as the Acadian Mountains were uplifted possibly as much as 4 km (Wrightstone 2010), the black and gray shale and siltstone of the Hamilton Group began accumulating in the foreland basin. Erosion of the mountains supplied these terrigenous sediments through the Catskill delta complex on the eastern margin of the basin (Schwietering 1981). The Marcellus Shale formed to the west in a sediment-starved, anoxic trough that paralleled the mountain chain. Large dust storms may have deposited aeolian quartz silt in the Marcellus (Mazzullo 1972, Prave et al.

1996, Wrightstone 2010).

During Marcellus time, the Appalachian Basin was located somewhere between 15° and

30° South latitude; however the exact location is still heavily debated. The subtropical warmth and increased sunlight in this location enhanced the growth of marine phytoplankton

(Wrightstone 2010). This plankton settled to the bottom and formed an organic-rich detritus, which was preserved in to the anaerobic environment (Boyce 2010).

The shallowest portion of the basin was along the eastern margin nearest the Acadian

Mountains. As the basin filled, seven deltaic depositional environments are recognized in southeastern New York in the Mahantango Formation: 1) alluvial and delta-plain, 2) wave and tide-dominated beach facies, 3) channel-mouth bar facies, 4) outer delta-platform facies, 5) interdistributary bay facies, 6) delta-slope facies, and 7) the prodelta facies. Water depth may have ranged from extremely shallow (less than 100 feet) to very deep (greater than 300 feet).

The vertical successions of these environments reflect the progradation of the delta complex over a shallow marine shelf during times of tectonic quiescence (Mazzullo 1972).

15 Prave et al. (1996) offered another interpretation of the Mahantango. Depositional cyclicity may have resulted from repeated progradation and retreat of a straight, tide-influenced shoreline system onto a storm-dominated marine shelf. This interpretation is based on three depositional systems inferred from the Mahantango Formation. The three systems are as follows: 1) storm dominated offshore marine mudstone and fine sandstones, 2) storm dominated near shore marine shelf and shoreface sandstone, and 3) tide-dominated shoreline sandstone, pebbly sandstone and mudstone. They are systematically arranged in thickening- and coarsening-upward cycles capped by transgressively reworked lag deposits (Prave et al. 1996).

In western New York strata equivalent to the lowest Mahantango (Skaneateles

Formation) include limestone, calcareous shale, gray shale, and organic-rich shale (Lash and

Engelder 2011). The organic-rich shale has a well-log character similar to that of the Marcellus and, in fact, can be mistaken for the Marcellus (Lash and Engelder 2011). These shale and limestone were deposited in an offshore prodeltaic environment far removed from the eastern shoreline.

In Ettensohn’s (2009) depositional model, clay-sized particles transported from the eastern Acadian Mountains remained suspended, finally coming to rest at the bottom of the

Appalachian Basin, tens of miles from the shore, at depths up to 500-900 ft or more. At these depths the water would have been sufficiently anoxic to preserve any accumulation of organic matter. This coupled with the rain-shadow effect of the Acadian mountains would have led to sediment starvation and further preserved the organic matter (Alegeo and Schekler 1998,

Walker-Milani 2011).

16 An alternate model shows that the deepest portion of basin may have been as shallow as

300 ft. or less (Schwietering 1981, Prave et al 1996, Boyce 2010, Walker-Milani 2011). At depths this shallow, the warm water needed to be sufficiently stratified so that the oxygen-rich portion of the water column did not mix with the anoxic bottom (Boyce 2010). Uranium and thorium identified in spectral gamma-ray logs were incorporated into the sediments syndepositionally, rather than being introduced into the formation later (Ettensohn 2008, Bank

2009). Sediment starvation occurred on the shallow shelf of the western margin whereas the basin center to the east acted as a trap on the clastic sediment from the mountains (Walker-

Milani 2011).

Another interpretation states that the preservation of the organic content was controlled by three main factors: 1) the photosynthesis of phytoplankton, 2) bacterial decomposition, and 3) bulk sedimentation rate (Wrightstone 2010). In this interpretation, organic–rich units were not deposited under deep, permanently anoxic-sulfidic water columns, but rather in an extremely shallow (less than 100 feet) stratified water column that may have changed seasonally. Walker-

Milani (2011) attributed the deposition primarily to terrigenous sediment supply that was dependent upon sea level. Low sediment supply favored the preservation of organic sediment; whereas high supply diluted the organics. A thermocline was established and destroyed seasonally (Werne et al. 2002, Walker-Milani 2011). The influx of aeolian sediment from giant dust storms on the basin margins may have triggered widespread algal blooms that produced basin-wide anoxia. The sudden die-off of these algal blooms promoted bacterial processes that acted to keep the bottom water anoxic, long enough to accumulate organic carbon on the sea floor (Wrightstone 2010). This, combined with the seasonal stratification and mixing, would

17 have promoted the recycling of biochemical nutrients and played a key role in the preservation and accumulation of the organic matter (Sageman et al. 2003).

The Marcellus organic-rich units were deposited and accumulated in an area of maximum organic production, maximum preservation, and minimum non-aeolian siliclastic sediment influx

(Wrightstone 2010). Relative sea-level rise and fall directly influenced the accumulation of the organic matter. Sea level rise led to sediment starvation and increased organic carbon concentrations in the distal basin sediments (Murphy et al. 2000). Siliclastic-sediment starvation is associated with organic carbon enrichment, and the highest level of total organic carbon content is related to the greatest transgression of relative sea level. High concentrations of aeolian silt are also indicative of high amounts of organic matter. An increase in non-aeolian clastic sediment influx and water column mixing during relative sea-level fall acted to dilute the surface sediment and decrease the volume of organic matter in the sediment (Sageman et al.

2003).

Ettensohn (1985) and Lash (2007) concluded that the accumulation of the black shale represents an episode of rapid, tectonic subsidence below the pycnocline and was linked to great water depth instead of marine algae blooms. The Marcellus Shale is considered the basal formation of the second tecto-phase of the Acadian Orogeny. During this phase, the Avalonian plate docked onto the continental margin of Laurasia, and the basin began to subside at a rate faster than that of sedimentation (Ettensohn 1985, Lash 2007). As the initial pulse of plate docking slowed, basin subsidence began to decrease.

18 Sedimentation rates eventually overcame the rate of subsidence and the coarse clastics of the overlying Mahantango Formation prograded over the distal mud. The first deposits of grey shale and siltstone above the Marcellus Shale mark this shallowing event.

Although tectonism acted as the main control on Marcellus and Mahantango sedimentation, other processes influenced local deposition. Biological processes produced the limestone, which may be found interbedded within the shale and siltstone. Repeated small-scale tectonic pulses, reflecting the minor movement of the thrust systems, may account for the close interbedding of black shale, siltstone, and limestone (Ettensohn 2005).

Once tectonism ceased, there was a period of quiescence and relative sea level lowstand.

The warm, shallow, sediment-starved sea promoted the deposition of the Tully Limestone

(Schwietering 1981, Ettensohn 1985).

As plate docking continued and the basin continued to subside, a foreland bulge began to uplift on the western side of the basin (Ettensohn 2005, Lash 2007, Lash and Engelder 2009).

This NE-SW trending tectonic feature played a key role in the stratigraphic thickness of the

Marcellus Shale in southwest New York and northwest Pennsylvania. The bulge formed a topographic high that led to local erosion of the Union Springs and, later, the Oatka Creek

Member (Lash and Engelder 2009). In addition to the foreland bulge, basement structures such as the Rome Trough and various lineaments that strike across the basin may have influenced sea- floor highs and lows. These highs and lows caused local variations in the thickness of the shale or created isolated sub-basins within the foreland basin (Lash 2007).

Lash and Engelder (2009) used a sequence-stratigraphic approach to model Marcellus sedimentation. In the underlying Onondaga Limestone lies a maximum regressive surface that

19 serves as a lower sequence boundary. The Marcellus Shale was deposited under two sequences of sea level rise and fall. The first sequence of transgression and regression began during the final deposition of the Onondaga Limestone. Sea level rose, flooding the basin, and the Union

Springs Member was deposited. The maximum surface of transgression is marked by a thin, pyritiferous condensed section near the base of the Union Springs. Above the Union Spring, the

Cherry Valley Limestone member was deposited during a time of relative tectonic quiescence and marks the maximum regressive surface or second sequence boundary of deposition. In the next depositional sequence sea level began to rise once again and the Oatka Creek was deposited. The maximum transgressive surface is once again marked by a highly radioactive, condensed section near the base of the Oatka Creek. Sea level then began to regress, as can be seen in the wireline logs as a coarsening-upward succession of the Oatka Creek. The Stafford

Limestone above the Marcellus was deposited during this regression and marks the next maximum regressive surface and the end of the second stratigraphic sequence (Lash 2007, Lash and Engelder 2009).

20 AREA OF INVESTIGATION

The focus area for this study (Figure 5) encompasses portions of four states: New York,

Pennsylvania, Ohio, and West Virginia, all of which fall within the Appalachian Basin. The extent of the generated maps cover approximately 83,000 square miles (214,000 sq-km)

(Wrightstone 2011). This area was chosen for two main reasons, (1) it contains a wealth of subsurface data in the form of gamma-ray and bulk-density logs and (2) it falls within the core area of Marcellus exploration and development.

Figure 5. Study area showing the Marcellus Shale (blue) in New York, Pennsylvania, Ohio, and West Virginia with the Devonian outcrops (grey) to the east.

22 METHODS OF INVESTIGATION

In this study, the Middle Devonian Marcellus Shale was analyzed to determine the extent of organic-rich rocks within the basin using criteria directly applicable to industry. The largest portion of available data came from geophysical logs utilizing gamma-ray and bulk-density curves. Gamma-ray and Density log data from 498 well logs in WV, OH, NY, and PA were utilized for this study. Where available, neutron and resistivity log data were also analyzed to better confirm the nature of each sedimentary package and to aid in correlation. Well logs from this study were primarily taken from proprietary data obtained from TGS-NOPEC Geophysical

Company. Where there are gaps in well-log data obtained from TGS, additional wells were obtained from the Oil and Gas Division of the Pennsylvania Geologic Survey and the West

Virginia Geologic and Economic Survey.

The first step, and the most important to this work, involved gathering and correlating the well logs throughout the study area. Figure 6 is a map of the well locations which were used for analysis. The Marcellus Shale and its members are the main stratigraphic units of interest and were correlated across the study area using the closed-loop method. The largest portion of the data was correlated using the gamma-ray logs. All of the log data was loaded into

GeoGraphixTM software, and correlations were made using the GeoAtlas module and XSection module of the program. Due to the variability of age among the well logs, an attempt was made to normalize curves which appeared to be too high or too low in API units from the standard type curves.

Figure 6. Location of data control points from digitized well logs (green) and rastor logs (purple) used in this study.

Formation tops, formation bottoms, and net thickness were determined for each organic- rich member. The author chose a cut-off value of greater than 160 API units between organic- rich shale (>160 API) and undifferentiated gray shale (100-160 API). The author chose a cut-off value of 100 API units to represent “clean strata” such as limestone and some non-organic shale.

Other API cut-offs were used to determine the varying amounts of radioactivity within the

Marcellus, including 180 API, 200 API, 250 API, and 300 API units, to determine which criterion is most effective in the interpretation of total organic content in the Marcellus Shale. In addition bulk-density cut-offs were also used: less than 2.5 G/cm3, less than 2.4 G/cm3, and finally less than 2.35 G/cm3. Thickness data was used to generate cross plots of the gamma-ray

API data vs the bulk-density data for a comparative analysis.

A series of isopach maps for each lithologic unit identified in the Marcellus shale were generated. Net thickness organic-rich shale maps were also created for the Marcellus. An isopach map of the gross thickness of the Mahantango Formation was also generated.

Ten cross-sections were generated, eight dip lines (SW-NE) and two strike lines (NE-

SW) (Figure 7). The positioning of each cross section was controlled by the geometry of the isopach maps. This was done in order to generate sections both on-trend and off-trend, especially in the Marcellus Shale.

Figure 7. Index line map for the ten cross sections generated in this study with total well spots overlaid. See appendix for further information on each well used in the cross sections.

26 The datum used to correlate the section is the top of the Onondaga Limestone. This datum is easily correlated and regionally widespread throughout the study area, except in the far west and eastern extents of the study area were the Onondaga Limestone pinches out.

A sequence stratigraphic analysis using the transgressive – regressive model (Catuneanu

2006, Embry 2009) of sedimentation was applied to the Marcellus in this study. Peaks in the gamma-ray represent maximum flooding surfaces whereas troughs represent maximum regressive surfaces. A sequence begins at the first maximum regressive surface and extends upward past the maximum flooding surface to the next maximum regressive surface higher in the section. Each surface was correlated across the study area, and isopach maps showing the thickness of each sequence and systems tract were generated. Systems tracts are taken to represent a single time slice of the local stratigraphy (Catuneanu 2006).

27 ISOPACH MAPS

MARCELLUS SHALE ISOPACH MAP

Figure 8 is the isopach map for the Marcellus Shale within the study area, and a number of features are apparent from its distribution. The body of the formation strikes N45°E with the depocenter [axis of greatest thickness approximately 220 feet (67 m) thick], occurring near the study’s eastern edge. From this thick, the shale thins to the west to a feather edge (10 feet or 3 m) in northwestern Pennsylvania and southwestern New York. Likewise, in West Virginia the thickest portion of the shale is approximately 150 feet (45.7 m) to the east, and thins to less than

20 feet in western West Virginia and southeastern Ohio. The strike of the Marcellus Shale depocenter seems to be shifted slightly eastward from the main strike of the Rome Trough, which is thought to have exerted a direct influence on basin deposition (Ettensohn 2005 (figure

9)).

28 1 2

Figure 8. Isopach of the Marcellus Shale where the red-dashed line represents the main body of deposition, the red solid lines represent cross strike thickness features, and the yellow-dashed line represents the trend of the thinning. Contour interval is 10 feet.

Figure 9. Map depicting the growth faults of the Rome Trough (From Ettensohn 2006).

30 Within the Marcellus Shale there appear to be 5 cross-strike thickening trends that are lobe-shaped (represented by solid red lines on Figure 8). The first occurs in northeastern

Pennsylvania and strikes due north into southern New York. The thickest portion of this cross- strike feature is approximately 190 feet (58 m) in Bradford County PA, to 120 feet (36.6 m) in

Tioga and Broome Counties, NY. The second cross-strike thickness trend is located in northeastern Pennsylvania and strikes northwest into southwestern New York. The thickest portion of this cross-strike feature is approximately 200 feet (87.8 m) in Lycoming County, PA, and thins to less than 60 feet (18.3 m) in Allegany and Steuben County NY. The third occurs in central Pennsylvania and strikes northwest into southwestern New York. The thickest portion of this cross-strike feature is approximately 180 feet (54.9 m) in Centre County, PA, to less than 40 feet (12.2 m) in Chautauqua County, NY. The fourth occurs in southwestern Pennsylvania and strikes northwest into eastern Ohio. The thickest portion of this cross-strike feature is approximately 140 feet (42.7 m) in Fayette County, PA, to less than 60 feet (12.2 m) in Jefferson

County OH. The fifth occurs in central West Virginia and strikes west-northwest into southeastern Ohio. The thickest portion of this cross-strike feature is approximately 150 feet

(45.7 m) in Webster County, WV, to less than 20 feet (6 m) in Meigs County, OH. These thickness trends coincide with lineaments (Fig. 10) that cross the strike of the basin and show a left lateral fault movement (Zagorski and Emery 2010).

Figure 10. Map of major basement faults and lineaments (Modified from Southworth 1891, Zagorski et al 2012).

In addition to the cross-strike thickening, there is a single line of major thinning (yellow dashed line on Figure 8). The thinning occurs in southwestern New York and trends into northwestern Pennsylvania and eastern Ohio where it wraps into southern West Virginia and continues to the southwest. The thickness of the Marcellus in this area ranges from 20 feet (6 m) at maximum to 10 feet (3 m) at minimum. In Grant and Pocahontas Counties, WV, there is an isolated area of thickening to about 160-220 feet (48.8 – 67.1 m).

UNION SPRINGS MEMBER ISOPACH

The Union Springs (Figure 11) is the lowest member of the Marcellus Shale. It is generally identified by the Cherry Valley Limestone capping the formation. It also displays a higher gamma-ray reading than the Oatka Creek above. The body of the formation strikes

N45°E with the thickest portion, approximately 120 feet (36.6 m) thick, occurring near the study’s eastern edge in northeastern Pennsylvania in Erie County. From this thick, the shale pinches out in northwestern Pennsylvania. The Union Springs also pinches out in New York, along the study's northern edge. Likewise, in West Virginia the thickest portion of the shale is approximately 60 feet (18.3 m) to the southeast, and thins to less than 10 feet in western West

Virginia and southeastern Ohio. Another pinch-out occurs in West Virginia, and encompasses much of Mason, Cabell, Putnam, and Kanawha counties.

33 1

Figure 11. Isopach Map of the Union Springs member of the Marcellus Shale. Contour interval is 10 feet.

34 CHERRY VALLEY MEMBER ISOPACH

The Cherry Valley (Figure 12) is the middle limestone member of the Marcellus Shale.

The Cherry Valley Limestone is easily identified on logs by a gamma-ray reading of less than

100 API. The body of the formation strikes N45°E and wraps due east in the northern portion of

PA. The thickest portion, approximately 12 feet (3.66 m) thick, occurring near the study’s western edge in southwestern Pennsylvania in Beaver County and another thick occurs in north central Pennsylvania in Potter County. In West Virginia the thickest portion of the limestone is approximately 8 feet (2.44 m) throughout the central portion of the state. The limestone thins both to the eastern and western portion of the study area in West Virginia, Pennsylvania, and

New York. The Cherry Valley also pinches out in New York along the study's northern edge.

Figure 12. Isopach map of the Cherry Valley Limestone of the Marcellus Shale. Contour interval is 10 feet.

36 OATKA CREEK MEMBER ISOPACH

Figure 13 is the isopach map generated for the Oatka Creek, which is the upper member of the Marcellus Shale. The depocenter of the formation strikes N45E with the thickest portion approximately 100 feet (30.5 m) thick near the study's northeastern edge in Pennsylvania and

New York. From this thick the unit thins to less than 10 feet (3 m) in western Pennsylvania and eastern Ohio. In West Virginia, the thickest portion of the shale is approximately 60 feet (18.3 m) in southeastern West Virginia. The shale thins to less than 20 feet to the west in southeastern

Ohio. The thinning of the Oatka Creek follows a similar trend to that of the Union Springs member.

This shale member displays 3 cross-strike features that are once again lobe-shaped. The first cross-strike feature occurs in southern New York and strikes to the northwest. Due to the strike, this feature correlates to the Lawrenceville–Altica Lineament. The lobe encompasses most of

Steuben County, NY, and is over 20 feet (6.1 m) thick as it extends into Ontario and Livingston

Counties, NY, where it thins to less than 10 feet (3 m). The second feature occurs in central

Pennsylvania and strikes northwest extending into southwestern New York; it correlates with the

Tyronne-Mount Union Lineament. The thickest portion of the feature is approximately 30 feet

(9.1 m) thick in central PA and thins to less than 20 feet (6.1 m) in northwestern PA. The third feature occurs in northern West Virginia and correlates with the Parsons and Highland County

Lineaments. The thickest portion of the feature is approximately 40 feet (12.19 m) thick in northern West Virginia and thins to less than 20 feet (6.1 m) in southeastern Ohio.

37 1

Figure 13. Isopach Map of the Oatka Creek Member of the Marcellus Shale. The red- dashed line represents the main body of the member and the red solid lines represent cross-strike features. Contour interval is 10 feet.

38 MAHANTANGO FORMATION ISOPACH

Figure 14 is the isopach map for the Mahantango Formation. The body of the formation strikes N45E. The thickest portion of the formation is 1600 feet (487.7 m) in northeastern

Pennsylvania and south-central New York with the very thickest portion located in Broome

County, NY. As the formation is traced to the west the thickness thins to less than 100 feet (30.5 m).

One cross-strike thickness features are apparent from the isopach map. It occurs in south central New York and strikes due north. This feature correlates loosely with the Lawerenceville-

Altica Lineament, and ranges in thickness from 1300 feet (396.2 m) to less than 1200 feet (365.8 m).

39 1

Figure 14. Isopach map of the Mahantango Shale the red-dashed line represents the main body of the formation.

40 STRATIGRAPHIC CROSS SECTIONS

Within the study area 8 dip lines and 2 strike lines were correlated (Fig 15). Dip section

A-A’ crosses southern New York and begins in Chautauqua County, NY and ends in Tioga

County, NY. Section B-B’ crosses northern Pennsylvania and begins in Erie County, PA and ends in Clinton County, PA. Section C-C’ crosses from eastern Ohio into central Pennsylvania and begins in Trumbull County, OH and ends in Cambria County, PA. Section D-D’ crosses from eastern Ohio into south central Pennsylvania and begins in Jefferson County, PA and ends in Somerset County, PA. Section E-E’ crosses from eastern Ohio to eastern West Virginia and begins in Monroe County, OH and ends in Tucker County, WV. Section F-F’ crosses from south eastern Ohio into central West Virginia and begins in Athens County Ohio and ends in Randolph

County, WV. Section G-G’ crosses from southeastern Ohio into southern West Virginia and begins in Gallia County, OH and ends in Greenbrier County, WV. Section H-H’ crosses from southwestern West Virginia to south central West Virginia and begins in Cabell County, WV and ends in McDowell County, WV.

Strike line I-I’ extends from New York southward into Ohio and ends in southwestern

West Virginia. The line begins in Erie County NY and ends in Cabell County, West Virginia.

Strike line J-J’ extends from New York southward into Pennsylvania and ends in south-central

West Virginia. The line begins in Broome County, NY and ends in Mingo County, WV.

Figure 15. Regional map showing the locations of all cross sections within the study area.

42 A-A’ CROSS SECTION